The Institute for Safe Medication Practices (ISMP) has joined forces with CVS and Walgreens pharmacies representing 20 percent of US pharmacies-and Outcome Engineering-recognized leaders in risk assessment modeling-to design risk models for high-alert medications dispensed from ambulatory pharmacies. Data from available sources-including error databases, scientific literature, and information gleaned from medication safety experts and focus groups-will be used to identify a preliminary list of medications dispensed from ambulatory pharmacies that carry a high risk of causing serious injury or death to a patient if they are involved in a medication error. Comprehensive risk models for these "high-alert" medications will be developed using four tools: process mapping, control system mapping, modified failure mode and effects analysis, and socio-technical probabilistic risk assessment (ST-PRA). Focus groups comprising pharmacists randomly selected from a convenience sample of pharmacies will provide input into the construction of fault trees that will be used to create the risk models. Focus groups and other reliable data sources will be used to estimate the frequency of failures within these processes, the severity of error outcomes, and dispensing frequency. For each fault tree, engineering software (Relex) will identify and rank [unreadable] the probability of all the failure combinations leading up to a harmful error. Validation of the resulting risk models will be accomplished through direct observation and interviews conducted at a stratified, random sample of six pharmacies, excluding stores involved in the initial model-building process. From the validated risk models, a subset of high-alert medications with the greatest probability of causing the greatest amount of harm will be determined and targeted for the identification of risk reduction strategies via an expert advisory panel. Knowing the medications with the highest probability of causing the greatest amount of harm, and the behaviors and system characteristics that are most likely to lead to errors with these medications, ambulatory pharmacies will be able to use the models to assess their risk and adjust their processes accordingly, thereby channeling precious resources into interventions most likely to impact patient safety. [unreadable] [unreadable] [unreadable]